Increase by Metyrapone Is Prevented by Dexamethasone Pretreatment

Journul of' Neuroendocrinolo~~,Vol. 2, No. 6 @ 1990 Oxford University Press, 0953-8194/90 $3.00

Characterization of Glucocorticoid Binding Capacity in Human Mononuclear Lymphocytes: Increase by Metyrapone is Prevented by Dexamethasone Pretreatment

Rainer Rupprecht, Johannes Kornhuber, Norbert Wodarz, Claudia Gobel, Johannes Lugauer, Christian Sinzger, Helmut Beckmann, Peter Riederer and Otto Albrecht Mullert Psychiatrische Klinik, University of Wurrburg, Wurzburg, FRG t Medizinische Klinik Innenstadt, Universlty of Munich, Munich, FRG

Key words. glucocorticoid binding, metyrapone, dexamethasone, lymphocytes, man

Abstract

Autoregulation of receptor systems by their own ligands is a well established biological phenomenon. While down-regulation of the glucocorticoid binding capacity by glucocorticoids has been shown in animals and humans, data on up-regulation processes in humans are lacking. To further explore glucocorticoid receptor plasticity in relation to endogenous ligands, glucocorticoid binding parameters were assessed in 15 healthy controls before and after oral administration of 1.5 g metyrapone with and without dexamethasone pretreatment. Administration of metyrapone resulted in blockade of the feedback of the hypothalamic-pituitary- adrenal system as shown by the rise in adrenocorticotropin levels, while pretreatment with 1 mg dexamethasone completely suppressed adrenocorticotropin concentrations. Glucocorticoid binding sites per lymphocyte exhibited an increase of 63% following metyrapone administration, which was prevented by dexamethasone pretreatment. Comparison of morning and afternoon glucocorticoid binding sites per cell in 11 healthy volunteers further revealed a diurnal rhythm of glucocorticoid receptor sites. These data suggest that human lymphocyte glucocorticoid receptors are under autoregulatory control.

Glucocorticoid receptors (GR) have been identified in numerous achieved by metyrapone (MET) administration in healthy volun- mammalian and human tissues including brain (1 -3), liver (4), teers mimicking the effect of adrenalectomy and whether this can lung (51, skin (6, 7),and peripheral blood cells (8-10). Although be rcvcrsed by exogenous glucocorticoids. several lines of evidence suggest an immunosuppressive effect of glucocorticoids with therapeutic implications (1 I), the exact mcch- anism of glucocorticoid action in relation to immune function remains unsolved. Glucocorticoids have been shown both to Results inhibit lymphocyte proliferation (1 1) and to inducc redistribution of circulating lcukocytcs to thc bone marrow (1 I). Phurmucologicul churactrristics qf' the GR Exposure to higher doses of glucocorticoids can down-regulate Linear regression analysis was performed on all saturation experi- the GR in animals (12) and humans (13, 14). In addition, ments; a representative Scatchard plot is shown in Fig. 1, Satura- investigations on GR pharmacology in human leukocytes at tion experiments clearly revealed only one binding site. The Hill different times of day (10, 15, 16) or in adrenocortical disorders coefficient was close to unity (1 .OOfO.OOl), indicating an absence (10, 17, 18) have provided conflicting evidence for a diurnal of cooperativity of the binding site. rhythm of GR (15) and dccrcascd GR numbers in adrenal Specific binding of 6 nM [3H]dexamethasone (['HIDEX) was insufficiency (18) or Cushing's syndrome (19), which others were completely displaced by the pure glucocorticoid agonist RU 28362 unable to confirm (10, 17). with a Ki value of 7.8f2.1 nM, by DEX with a Ki value of Down-regulation of GR by glucocorticoids has been relatively 13.4 & 1.7 nM, and by cortisol with a K, value of 117 f 3 1 nM, easy to study in either experimental animals or man, while data respectively, under equilibrium conditions in three different on up-regulation phenomena are available only from animals experiments. after adrenalectomy (20). Thc present study was designed to There was a linear relationship between absolute CR and explore whether an increase in GR binding capacity can be lymphocyte number in the assay.

~ Correspondence 10: R. Rupprecht, Max-Planck-Institute of Psychiatry, Kraepelinstr. 2- 10, 8000 Munchen 40. FRG 804 Glucocorticoid binding capacity in human lymphocytes SCATCHARD PLOT GLUCOCORTICOID RECEPTOR SITES PER CELL

B/F -3 6000 3 1 5000

4000

3000

\ \ 2000 1000 n- 0 / Baseline MET MET+DEX 0 5 10 15 20 Bound [M]-’*/lO ’ cells GLUCOCORTICOID RECEPTOR REGULATION FIG.1, Scatchard plot of a representative saturation experiment of specific FIG.3. Mean f SE plasma glucocorticoid receptor (GR) binding sites per [3H]dexamethasone binding to human mononuclear leukocytes. Specific- cell at 1600 h on three consecutive days in 15 healthy controls under ally bound/free [3H]dexamethasone is plotted versus specifically bound baseline, metyrapone (MET) and metyrapone plus dexamethasone [3H]dexamethasone. (MET + DEX) pretreated conditions. GR sites per cell are significantly increased by MET administration (t= 3.98, df= 14, P

ACTH (pmol/L) FIG.4. Meanf SE K, values at 1600 h on three consecutive days in 15 healthy controls under baseline, metyrapone (MET) and metyrapone plus 35 dexamethasone (MET + DEX) pretreated conditions.

30 subsequent replacement by DEX did not significantly alter the 25 K, (Fig. 4). 20 No relationship was found between age, sex, body weight, ACTH, or baseline cortisol levels (223.7k75.6 nmol/L) and GR 15 binding parameters under the conditions studied. 10 Discussion 5 Although previous investigations of a possible diurnal rhythm of 0 GR binding parameters have yielded contradictory results (10, Baseline MET MET*DEX 15, 16), the present study demonstrated clear diurnal rhythm of FIG.2. Mean SE plasma adrenocorticotropin (ACTH) concentrations GR sites with no apparent change in GR affinity. Interestingly, at 1600 h on three consecutive days in 15 healthy controls under baseline, we did not observe a significant correlation between GR sites metyrapone (MET) and metyrapone plus dexamethasone (MET + DEX) and ACTH or cortisol levels, suggesting that plasma concentra- pretreated conditions. ACTH levels are significantly increased by MET administration (t=4.5, df= 14, P <0.001). This increase is prevented by tions of endogenous glucocorticoids at the time of the GR assay DEX pretreatment. do not control the GR sites. Estimation of cortisol profiles several Glucocorticoid binding capacity in human lymphocytes 805 hours prior to that of GR sites however might clarify the France). Rotiszint 22 was obtained from Roth (Karlsruhe, West Ger- relationship of GR sites and endogenous glucocorticoids. many). The incubation buffer consisted of PBS with 5 mM D-glucose added. Stock solutions of unlabelled steroids (50 pM) were prepared in We demonstrated an increase in GR activity binding 7 h incubation buffer containing 5% ethanol. following MET pretreatment. At this time ACTH levels had also increased dramatically due to the lack of feedback action of 1113- Preparation of cells hydroxylated endogenous glucocorticoids. This increase could be A mononuclear cell fraction was prepared by sodium metrizoate-Ficoll prevented by prior administration of the synthetic glucocorticoid density gradient centrifugation (26). Cells were washed two times in PBS for 10 min, incubated for 60 min at 37 "C, and then washed again to DEX. Since glucocorticoids probably need several hours to exert allow dissociation of endogenous hormone. Endogenous cortisol levels a notable influence on GR regulation, this might explain the lack after the washing procedure were below the detection limit of the assay. of correlations between GR sites and plasma cortisol concentra- The final concentration of cells was determined using a Coulter Counter tions at the time of the GR assay. (Model S5; Coulter Electronics Ltd, England). Viability of cells exceeded 95%, as judged from their ability to exclude trypan blue. Contamination Our results in humans accord with those of animal studies and by erythrocytes was less than 10%. contamination by granulocytes and recent in vitro investigations at the protein and DNA level. A monocytes was less than 8% and did not differ between the three test rise in GR sites in the rat brain has been observed following days. adrenalectomy (20), while exposure to glucocorticoids results in Binding assay a down-regulation of GR sites of various cell lines (13, 14, 21). Binding experiments were carried out at 37 'C in plastic microtitre plates While glucocorticoids reduce GR mRNA (22, 23) and GR protein in a total volume of 0.25 mL. The displacing compound (10 pM unlabelled (24) within several hours following binding of GR to GR cDNA, DEX) was added immediately prior to the addition of [3H]DEX to receptor degradation may also be involved as an autoregulatory determine nonspecific binding. In saturation studies increasing concentra- post-translational control mechanism independent of GR protein tions of [3H]DEX from 1 to 40 nM were used. Saturation experiments were performed at equilibrium after 90 min incubation. After incubation synthesis (21, 25). bound ligand was separated from free ligand by rapid filtration through Although changes in GR binding do not necessarily reflect Scatron filters with a Titertek cell harvester by two 5-s washes with assay changes in receptor mass, since the results of binding studies are buffer at room temperature. The fillers were transferred into plastic vials, dependent on the availability of receptors for exchange with 5 mL of a toluene-based scintillation cocktail was added (Rotiszint 22) radiolabelled tracer, an exact estimation of GR sites after sufficient and they were monitored for tritium in a Beckman LS 1801 counter at about 54% efficiency. All samples were assayed in triplicate with a dissociation of endogenous glucocorticoids has been claimed (1 3). variation within a single experiment of less than 7%. The use of monoclonal antibodies and mRNA probes (24) may help to discriminate between transcriptional and post-transla- Hormone as.say.s tional phenomena and further our understanding of human ACTH was measured by a newly developed immunoradiometric assay supplied by the Nichols Institute (San Juan Capistrano, CA, USA), which physiology. does not require extraction procedures (27). A soluble sandwich complex is formed by a ['251]labelled monoclonal antibody directed against N-terminal ACTH and a biotin-coupled polyclonal antibody against C-terminal Materials and Methods ACTH. The sandwich complexes are bound by adding avidin-coated plastic beads. Unbound components are washed away and the radioactiv- Subjects ity bound to the solid phase is monitored in a gamma-counter. The lower Fifteen subjects (three men and twelve women) participated in the study detection limit was 1.5 pmol/L, and the intra- and interassay coefficients on a voluntary basis. Their mean age and body weight were 47.3k6.9 of variation were 3% and 6.8%, respectively. (+SD) years and 68.8 k9.9 kg, respectively. Exclusion criteria included Cortisol was measured by a direct RIA (28). The lower detection limit illness, intake of medication, alcohol or nicotine abuse and stressful life was 25 nmol/L, and the intra- and interassay coefficients of variation events prior to the investigation. Five women were tested during the were 5% and I YO, respectively. As there is cross-reactivity of most cortisol midluteal phase of the cycle and seven were postmenopausal. Diurnal antibodies with I I-deoxycortisol (about 15% in our assay), assessments variation of GR characteristics was assessed in an additional I1 healthy of cortisol values following MET application would not be accurate and volunteers (six men and five women) with a mean age of 35.2k 15.2 years were therefore excluded from further analysis. and a mean body weight of 67.2k8.9 kg. Data analysis Test protocol Preliminary estimates of binding parameters from saturation experiments All subjects were admitted to a sleep laboratory at least 1 h prior to drug were provided by the EBDA program (29). Final estimates of binding administration or blood sampling. The test protocol was approved by parameters were determined with a computerized non-linear, least-square the ethical committee of the University of Wiirzburg. For determination regression analysis (30). This weighted curve fitting program assumes of GR binding characteristics and hormone data 50 ml blood samples binding according to the law of mass action to independent classes of were collected at 1600 h into prechilled plastic tubes containing EDTA binding sites. on three consecutive days. On Day 1 baseline values were obtained. On The results are expressed as the meankSD, and as the meankSE in Day 2 the subjects were pretreated with 1.5 g MET at 0900 h administered Figs. 2 to 4. orally with milk to avoid severe gastric symptoms. One mg DEX was Data were analysed using the I-test for paired samples and Pearson's given orally at 2300 h. On Day 3 administration of 1.5 g MET was product moment correlation. All significance levels are two-tailed. performed as described for the day before. Twelve of the 15 subjects experienced a transient dizziness and a hot flush, which disappeared Accepted 12 June 1990 within 45 to 60 min. For the evaluation of diurnal GR binding parameters blood samples were obtained twice at 0800 h and 1600 h. References Chemicals I. Fuxe K, Wikstrom AC, Okret S, Agnati LF, HPrfstrand A, Yu ZY, [3H]DEX ([I ,2,4,6,7-3H]dexamethasone, specific activity 84 Ci/mmol) was Granhol L, Zoli M, Vale W, Gustafsson JA. (1985). Mapping of obtained from Amersham, UK. Unlabelled DEX, cortisol, phosphate- glucocorticoid receptor immunoreactive neuro.is in the rat tel- and buffered saline (PBS) and sodium metrizoate-Ficoll were obtained from diencephalon using a monoclonal antibody against the rat liver gluco- Sigma (St. Louis, MO, USA), and RU 28362 from Roussel-Uclaf (Paris, corticoid receptor. Endocrinology. 117: 1803-1812. 806 Glucocorticoid binding capacity in human lymphocytes 2. McEwen B, Chao H, Spencer R. Brinton R, Macisaac L, Harrelson clear leukocytes (PMN) and mononuclear cells (MN) in a normal A. (1987). Corticosteroid receptors in brain: relationship of receptors population. J Steroid Biochem. 25: 483-487. to effects in stress and aging. Ann NY Acad Sci. 512: 394-401. 17. Kontula K, Pelkonen R, Anderson L, Sivula A. (1980). Glucocort- 3. De Kloet ER, Ratka A, Reul JMHM, Sutanto W, van Eekelen JAM. icoid receptors in adrenocortical disorders. J Clin Endocrinol Metab. (1987). Corticosteroid receptor types in brain: regulation and putative 51: 654-657. function. Ann NY Acad Sci. 512: 351-361 18. Schlechte JA. Sherman BM. (I 982). Decreased glucocorticoid 4. Westphal HM, Moldenhauer G. Beato M. (1982). Monoclonal anti- receptor binding in adrenal insufficiency. J Clin Endocrinol Metah. 54: bodies to the rat liver glucocorticoid receptor. EMBO J. 1: 1467-1471. 145 149. 5. Ballard PL, Ballard RA. (1974). Cytoplasmic receptors for glucocort- 19. Davydova IV, Nikolaeva NY, Golilov PP. Kalinin AP. (1987). icoids in lung of the human fetus and neonate. J Clin Invest. 53: Glucocorticoid receptors of blood lymphocytes in Icenko-Cushing 477-486. disease. Probl Endokrinol (Mosk). 33: 28-31. 6. Berkovitz GD, Carter KM, Migeon CJ. Brown TR. (1988). Down- 20. Reul JMHM, van den Bosch R, de Kloet ER. (1987). Relative regulation of the glucocorticoid receptor by dexamethasone in cultured occupation of type I and type 11 corticosteroid receptor; in rat brain human skin fibroblasts: implications for the regulation of aromatase following stress and dexamethasone treatment: functional implications. activity. 3 Clin Endoerinol Metab. 66: 1029 -1036. J Endocrinol. I 15: 459- 467. 7. Oikarinen A, Oikarinen H, Meeker CA, Tan EML. Uitto K. (1987). 21. Svec F, Rudis M. (1981). Glucocorticoids regulate the glucocorticoid Glucocorticoid receptors in human skin fibroblasts: evidence for down- receptor in the AtT-20 cell. J Biol Chem. 256: 5984-5987. regulation of receptor by glucocorticoid hormone. Acta Derm Venereol 22. Okret S, Poellinger L, Dong Y, Gustafsson JA. (1986). Down- (Stockh). 67: 461-468. regulation of glucocorticoid receptor mRNA by glucocorticoid hor- 8. Lippman M, Barr R. (1977). Glucocorticoid receptors in purified mones and recognition by the receptor of a specific sequence within a subpopulations of human peripheral blood lymphocytes. J Immunol. receptor cDNA clone. Proc Natl Acad Sci USA. 83: 5899-5903. 118: 1977-1981. 23. Kalinyak JE, Grifin CA, Hamilton RW, Bradshaw JG, Perlman AJ, 9. Martins VR, Birolli MIA, Duarte AJS, Brentani MM. (1987). Gluco- Hoffman AR. (1989). Development and hormonal regulation of glucocorticoid receptors in subpopulations of human lymphocytes defined corticoid receptor messenger RNA in the rat. J CLin Invest. 84: 1843- by monoclonal antibodies. Cell Immunol. 105: 443-446. 1848. 10. Junker K. (1983). Glucocorticoid receptors of human mononuclear 24. Dong Y, Poellinger L, Gustafsson JA, Okret S. (1988). Regulation leukocytes in vitro. J Clin Endocrinol Metah. 57: 506-512. of glucocorticoid receptor expression: evidence for transcriptional and 11. Cupps TR, Fauci AS. (1982). Corticosteroid-mediated immunoregul- posttranslational mechanisms. Mol Endocrinol. 2: I256 - 1264. ation in man. lmmunol Rev. 65: 133- 155. 25. Mclntyre WR, Samuels HH. (1985). Triamcinolone acetonide regu- 12. Reul JMHM, van den Bosch FR, de Kloet ER. (1987). Differential lates glucocorticoid-receptor levels by decreasing the half-life of the response of type I and type 11 corticosteroid receptors to changes in activated nuclear-receptor form. J Biol Chem. 260: 418-427. plasma steroid level and circadian rhythmicity. Neuroendocrinology. 45: 26. Boyum A. (1986). Separation of leukocytes from blood and bone 407-41 2. marrow. Scand J Clin Lab Invest. 21: 97-107. 13. Schlechte JA, Ginsberg BH, Sherman BM. (1982). Regulation of the 27. Raff H, Findling JW. (1989). A new immunoradiometric assay for glucocorticoid receptor in human lymphocytes. J Steroid Biochem. 16: corticotropin evaluated in normal subjects and in patients with Cush- 69-74, ing's syndrome. Clin Chem. 35: 596-600. 14. Shipman GF, Bloomfield CD, Gajil-Peczalska K, Munck AU, Smith 28. Stalla GK, Giesemann G, Miiller OA, Wood WG, Scriba PC. (1981). KA. (1983). Glucocorticoids and lymphocytes. 111. Effects of glucocort- The development of a direct homologous radioimmunoassay for serum icoid administration on lymphocyte glucocorticoid receptors. Blood. cortisol. J Clin Chem Clin Biochem. 19: 427-434. 61: 1086-1090. 29. McPherson GA. (1983). A practical computer-based approach to the 15. Schlaghecke R, Kley HK. (1986). Circadian and seasonal variations analysis of radioligand binding experiments. Comput Methods Programs of glucocorticoid receptors in normal human lymphocytes. Steroids. Biomed. 17: 107-114. 47: 287- 294. 30. Munson PJ, Rodbard D. (1980). LIGAND: a versatile computerized 16. Doe RP, Goldman PM, Severson SP, Hruby HM. (1986). Circadian approach for characterization of ligand-binding systems. Anal Biochem. variation of cytosol glucocorticoid receptors in human polymorphonu- 107: 220-239.